English Deutsch
Новости
Мир антропологии

Black Neanderthals, black blue-eyed Cro-Magnons

Anthropologists have been investing significant effort in reconstructing the appearance of ancient humans, but who and why would be interested in the result? What is the use for knowing archaic human phenotype? We have an answer: from the very beginning, the reconstruction methods, in addition to merely satisfying scientists’ curiosity, had a very practical application – the forensic examination. In particular, anthropologists helped in putting names on unidentified remains.

Homo neanderthalensis. Credit: Roman Evseev
Homo neanderthalensis. Credit: Roman Evseev

For a long time geneticists have been working side by side with classic anthropologists in the field of forensic science: they determine the identity of the criminal by samples of blood, saliva, semen or even by the traces of sweat on objects the perpetrator has touched. This is possible if the suspect’s genetic “autograph” is present in the police database. What if it is not? In this case, we can use the geno-geography to find out the probable region of the person’s origin. However, this can also be a problem due to poor DNA preservation. The reconstruction of the genome owner’s appearance is an answer. Person’s hair, eyes and skin color will be the easiest things to determine. The genetics of pigmentation has been developing for a long period. By now, specialists identified dozens of genes affecting the synthesis and distribution of pigment in various human tissues.

The Forensic Science International: Genetics journal published an article on the development the DNA-based methods of prediction of human pigmentation [1]. The techniques were designed for searching for criminals, acquitting innocent suspects, and for identifying victims of accidents and natural disasters. A few years ago the authors designed a system called HIrisPlex (H for hair). From its name it is easy to figure out that it was designed to deduce eye and hair color from DNA sequence. Predicting skin pigmentation was more challenging due to high number of genetic variants associated with the variance in eye and hair color. Therefore, it was easier for geneticists to study a relatively homogeneous group of people which has evolved within a particular continent. Human skin provides quite a different, more complex, planet-wide story. We know, for example, that in the course of evolution Europeans and Asians became light-skinned at least partially independently – their lack of skin pigmentation have different genetic bases.

Back in 2010 geneticists developed a version of their algorithm for predicting the skin color based just on three SNPs (SNP – single nucleotide polymorphism. It is a position of DNA sequence which can hold different “letters of genetic alphabet” in different individuals) [2]. An improved protocol was published a year later. It analyzed 7 SNPs predicting three types - "not pale", "not dark", or "not pale / not dark" (i.e. intermediate). The version released later took into account 17-18 SNPs and it predicted "dark/pale skin" with 67% probability.

Finally, in 2014, authors estimated the predictive ability of 59 SNPs and selected 10 of them, which produced good prediction [3]. However, the applicability of the product was low due to the limited number of tested individuals. In 2017, the authors extended their selection up to 2,025 individuals from around the world and tested 77 SNPs of 37 genes [4]. This resulted in a model that was able to place a tested genotype into one of five color categories the skin, from very pale to very dark ("Dark-Black"), with reasonably high probability. The technique was tested on 200 individuals from various human populations, with pretty decent result. For example, it managed to predict "Dark-Black" skin in 98% of cases.

After combining their successful algorithms, the authors developed a unified system of DNA testing HIrisPlex-S (S for skin). Seventeen SNPs used for predicting the skin color were combined with 24 SNPs "for eyes and hair" (19 of them also affecting skin pigmentation). The technique estimates the probabilities for 3 eye colors, 4 hair colorations and 5 skin types.

Figure 1. Global skin colour prediction with the HIrisPlex-S DNA test system.
								Source: https://www.fsigenetics.com/article/S1872-4973(18)30220-5/fulltext
Figure 1. Global skin colour prediction with the HIrisPlex-S DNA test system.
Source: https://www.fsigenetics.com/article/S1872-4973(18)30220-5/fulltext

For calibrating the system, genetic data of hundreds individuals from different parts of the world were included in the analysis.  Special focus was made on the samples of a "forensic" type: blood stains, saliva, semen, vaginal smears, and traces of touch on various objects.

The article pays special attention to the description of sample preparation procedure, as well as to the testing of technique reliability. For example, what happens if we mix DNA samples of two different persons? The tests demonstrated that the technique determined which sample was mixed, although not in every case. Will the technique distinguish human DNA from animal DNA? According to the tests, it will, except for the chimpanzees – our closest relatives are too similar to us. Will the results be reproduced in different laboratories? Yes, the tests in five independent laboratories were successful, producing correct results in 28 out of 30 samples.

In the process of testing the algorithm, the authors put the results of skin color prediction of over 1000 individuals on the world map (Figure 1). The resulted picture was similar to the observed geographical distribution of human pigmentation.

Illustrative example of the performance of the skin colour prediction model provided with the HIrisPlex-S DNA test system.
									Source: https://www.fsigenetics.com/article/S1872-4973(18)30220-5/fulltext
Illustrative example of the performance of the skin colour prediction model provided with the HIrisPlex-S DNA test system.
Source: https://www.fsigenetics.com/article/S1872-4973(18)30220-5/fulltext

Another illustration of the technique’s effectiveness is the correct prediction of the skin color of 35 arbitrarily selected individuals from America, Europe, Africa and Asia. Importantly, these individuals were not included in the database used in the calibration of the model. Figure 2 indicates high correlation of skin photographs with the results of the HIrisPlex-S method.

Authors emphasized that the technique demonstrated excellent results in samples of poor quality with very low content of DNA – as little as 63pg. The developers plan to further improve the method by including new SNPs into the procedure for better reliability of prediction.

The researchers drew attention to the fact that their program might be applied not only to forensic science, but also to other areas, no doubt including evolutionary biology and paleoanthropology.

Now we proceed to the most interesting part. Among other things, the developers published a free online version of the technique. To use the pigmentation calculator at https://hirisplex.erasmusmc.nl/ you can fill in the variants of 41 SNPs from the genome of interest and get a prediction!

We do have an appropriate genome, and even a few. The website of Max Planck Institute for Evolutionary Anthropology (https://bioinf.eva.mpg.de/jbrowse/) has freely accessible genetic sequences of ancient humans. We could not resist taking advantage of this opportunity and tested the genomes of two Neanderthals (from Denisova cave in Altai and from Vindija cave in Croatia), a Denisovan, and a 45 thousand years old Homo sapiens from Ust-Ishim (Siberia) in Hirisplex.

We are grateful to Dr. Susan Walsh, Ph.D., Assistant Professor, Biology Department, Forensic & Investigative Sciences Program for clarifying the details of how to select SNP calls for Hirisplex.

The result came as a surprise:

 

Eyes

Hair

Skin

Altai Neanderthal

Brown

Dark brown / black

Dark to black

Vindija Neanderthal

Brown

Dark brown / black

Dark to black

Denisovan

Brown

Dark brown / black

Dark to black

Ust-Ishim man

Blue

Dark brown

Dark to black

According to Hirisplex, and providing we did not make any mistakes, both Neanderthals, as well as the Denisovan, had very dark skin. The Ust-Ishim man’s skin turned out to be just as dark, though this ancient Homo sapiens stood out with his potentially blue eyes (with probability 0.678), due to a T in the SNP rs12896399 at locus LOC105370627.

Interestingly, the Cheddar man’s reconstructed portrait (Britain, 10 thousand years ago) had the same look – dark skin and blue eyes, as well as other European Mesolithic hunters, who lived much later – 7-8 thousand years ago (see the studies of 2014 [5] and 2015 [6]). This suggests that genetic trait responsible for blue eyes had spread among our ancestors long before the European inhabitants became light skinned.

The uniform blackness of all four archaic individuals had puzzled us so much, that we became concerned about software artifacts. As controls, we analyzed the genomes of a French and a Mbuti pygmy from the same website of Max Planck Institute for Evolutionary Anthropology. The results were as follows:

 

Eyes

Hair

Skin

French

Blue

Dark brown / black

Intermediate / pale

Mbuti

Brown

Dark brown / black

Dark to black

Even though we did not know the actual pigmentation of these French and Mbuti individuals, the results looked plausible. Yet, before spreading the word about black Neanderthals all over the world, we must make an important disclosure. HIrisPlex-S is a technique designed for testing modern humans. We have mentioned above that even within our species the evolution of the pigmentation went different ways in different populations. Both Neanderthals and Denisovans had been evolving in Eurasia for hundreds of thousands years separately from our ancestors, so they could acquire light skin, eyes or hair in their own way, as a result of mutations yet unknown to us. In this case, the technique would not work, because such unique Neanderthal’s SNPs were not represented in the HIrisPlex-S calculator.

It should be pointed out that the famous Neanderthal red hair, of which a lot has been written after the research in 2007 in which scientists had obtained DNA fragments from Neanderthal remains from Monti Lessini (Italy) and El Sidron (Spain). Note that the mutation in MC1R gene which, according to the scientists, had resulted in the Neanderthals having red-hair was not found in modern humans.

We did not find this mutation of red hairness in the Neanderthals from Denisova cave and Vindia. Very likely, the hair color of Neanderthals in Eurasia was variable like in modern humans.

Based on our limited amateur research, we can draw the following conclusions:

1. Taking into account only those SNPs that determine the color of skin, hair and eyes in modern humans and represented in HIrisPlex-S, the Neanderthals from Denisova cave and Vindja, as well as the Denisovian and the ancient Homo sapiens from Ust-Ishim were brunettes with very dark skin. The Ust-Ishim man potentially had blue eyes (although, the probability in favor of Ust’-Ishim’s blue eyes was not overwhelmingly high), while the eyes of other specimens tested were brown.

2. The common ancestor of Homo sapiens, Homo denisova and Homo neanderthalensis was a dark-skinned brunette with brown eyes. The Eurasian Homo sapiens have acquired their light skin independently of the Neanderthals, if the latter had ever been pale.

The flaws of our analysis and the future directions of research are the following:

We were limited to the set of SNPs represented in HIrisPlex-S. Additional studies of archaic genomes are necessary to identify the mutations absent in modern human population which could have influenced Neanderthals’ pigmentation.

Siberia is a cold place today, especially in winter. Was it any warmer 50 thousand years ago, or not, the question of whether Altai Neanderthals wore clothes remains open. Lack of clothing could be a factor of selection in favor of retaining the dark pigmentation in Neanderthals.

If the Neanderthals and the Denisovans of Altai did not wear clothes, could they possess genetically determined resistance to the winter cold? We can recall Konstantin Anisimov’s and Verona Konrad’s reports about cold-proof Papuans walking naked at +8°C [ref]. It is the Papuan genome that contains the highest proportion of the Denisovan DNA. We can speculate that archaic SNPs provide Papuans with tolerance to low temperature. Similarly, several researches demonstrated that not all populations of ancient humans in Europe used fire regularly [7].

The archaic genomes still hold many surprises for us. The availability of full-size genomes and of the tools for their analysis allows any enthusiast in genetics to make small discoveries, similar to amateur astronomers finding new comets and asteroids.

The results of testing HIrisPlex-S, in detail:

Altai Neandertal

Input data:

Predicted phenotype:

p-value
blue eye 0,338
intermediate eye 0,115
brown eye 0,547
blond hair 0,065
brown hair 0,652
red hair 0,001
black hair 0,283
light hair 0,13
dark hair 0,87
very pale skin 0
pale skin 0
intermediate skin 0
dark skin 0,04
dark to black skin 0,996

Vindija Neandertal

Input data:

Predicted phenotype:

p-value
blue eye 0,338
intermediate eye 0,115
brown eye 0,547
blond hair 0,065
brown hair 0,652
red hair 0,001
black hair 0,283
light hair 0,13
dark hair 0,87
very pale skin 0
pale skin 0
intermediate skin 0
dark skin 0,004
dark to black skin 0,996

Denisova

Input data:

Predicted phenotype:

p-value
blue eye 0,338
intermediate eye 0,115
brown eye 0,547
blond hair 0,086
brown hair 0,673
red hair 0,002
black hair 0,24
light hair 0,216
dark hair 0,784
very pale skin 0
pale skin 0
intermediate skin 0
dark skin 0,004
dark to black skin 0,996

Ust-Ishim Homo sapiens

Input data:

Predicted phenotype:

p-value
blue eye 0,678
intermediate eye 0,083
brown eye 0,239
blond hair 0,136
brown hair 0,619
red hair 0,002
black hair 0,243
light hair 0,352
dark hair 0,648
very pale skin 0
pale skin 0
intermediate skin 0,003
dark skin 0,096
dark to black skin 0,902

French

Input data:

Predicted phenotype:

p-value
blue eye 0,926
intermediate eye 0,057
brown eye 0,017
blond hair 0,254
brown hair 0,638
red hair 0,004
black hair 0,104
light hair 0,675
dark hair 0,325
very pale skin 0,024
pale skin 0,402
intermediate skin 0,572
dark skin 0,001
dark to black skin 0,001

Mbuti

Input data:

Predicted phenotype:

p-value
blue eye 0,338
intermediate eye 0,115
brown eye 0,547
blond hair 0,065
brown hair 0,652
red hair 0,001
black hair 0,283
light hair 0,13
dark hair 0,87
very pale skin 0
pale skin 0
intermediate skin 0,001
dark skin 0,018
dark to black skin 0,982

Translation: Lubov Zhuk

Source (in Russian): Портал XX2 Век

References

  1. L. Chaitanya, K. Breslin, S. Zu?iga, L. Wirken, E. Po?piech, M. Kukla-Bartoszek, T. Sijen, P. de Knijff, F. Liu, W. Branicki, M. Kayser1, S. Walsh. The HIrisPlex-S system for eye, hair and skin colour prediction from DNA: Introduction and forensic developmental validation. J Foren Sci, 2018. 35: p. 123-135.
  2. R.K. Valenzuela, M.S. Henderson, M.H. Walsh, N.A. Garrison, J.T. Kelch, O. Cohen-Barak, et al., Predicting Phenotype from Genotype: Normal Pigmentation*. J Foren Sci, 2010. 55(2): p. 315-322.
  3. O. Maro?as, C. Phillips, J. S?chtig, A. Gomez-Tato, R. Cruz, J. Alvarez-Dios, et al., Development of a forensic skin colour predictive test. Foren Sci Int Genet, 2014. 13: p. 34-44.
  4. S. Walsh, L. Chaitanya, K. Breslin, C. Muralidharan, A. Bronikowska, E. Pospiech, et al., Global skin colour prediction from DNA. Hum Genet, 2017. 136(7): p. 847-863.
  5. I?igo Olalde et al. Derived immune and ancestral pigmentation alleles in a 7,000-year-old Mesolithic European. Nature, 2014. 507: p. 225–228.
  6. Iain Mathieson et al. Genome-wide patterns of selection in 230 ancient Eurasians. Nature, 2015. 528: p. 499–503.
  7. Wil Roebroeks, Paola Villa. On the earliest evidence for habitual use of fire in Europe. PNAS March 14, 2011.



06 октября - Ученые Против Мифов в Москве

Интересно

Как заметил Ж.Ж. Руссо, «обезьяна подражает человеку, которого она бо­ится, и не подражает презираемым ею животным; она находит правильным то, что делает высшее по сравнению с ней существо».

Руссо, Ж.Ж. Педагогические сочинения: В 2-х томах М., 1981. Том 1. С. 230. Цитата предоставлена Викентьевым И.Л.

Catalog gominid Antropogenez.RU